<html><head></head><body>I've asked before and not had any luck getting someone to try this:<br>
<br>
On the server you're about to reboot, "killall glusterfsd" and let me know if you still see the same problem at the client. <br><br><div class="gmail_quote">On April 9, 2014 1:19:43 AM PDT, Fabio Rosati <fabio.rosati@geminformatica.it> wrote:<blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
<pre class="k9mail">Hi Paul,<br /><br />you're not alone. I get the same issue after rebooting a brick belonging to a 2 x 2 volume and the same is true for João P. and Nick M. (added in cc).<br /><br />[root@networker ~]# gluster volume info gv_pri<br /> <br />Volume Name: gv_pri<br />Type: Distributed-Replicate<br />Volume ID: 3d91b91e-4d72-484f-8655-e5ed8d38bb28<br />Status: Started<br />Number of Bricks: 2 x 2 = 4<br />Transport-type: tcp<br />Bricks:<br />Brick1: nw1glus.gem.local:/glustexp/pri1/brick<br />Brick2: nw2glus.gem.local:/glustexp/pri1/brick<br />Brick3: nw3glus.gem.local:/glustexp/pri2/brick<br />Brick4: nw4glus.gem.local:/glustexp/pri2/brick<br />Options Reconfigured:<br />storage.owner-gid: 107<br />storage.owner-uid: 107<br />server.allow-insecure: on<br />network.remote-dio: on<br />performance.write-behind-window-size: 16MB<br />performance.cache-size: 128MB<br /><br /><br />I hope someone will address this problem in the near future since not being able to
shutdown a server hosting a brick is a big limitation.<br />It seems someone solved the problem using cgroups: <a href="http://www.gluster.org/author/andrew-lau">http://www.gluster.org/author/andrew-lau</a>/<br />Anyway, I think it's not easy to implement because cgroups is already configured and in use for libvirt, if I had a test environment and some spare time I would have tried.<br /><br /><br />Regards,<br />Fabio Rosati <br /><br /><br />----- Messaggio originale -----<br />Da: "Paul Penev" <ppquant@gmail.com><br />A: Gluster-users@gluster.org<br />Inviato: Domenica, 6 aprile 2014 17:52:53<br />Oggetto: [Gluster-users] libgfapi failover problem on replica bricks<br /><br />Hello,<br /><br />I'm having an issue with rebooting bricks holding images for live KVM<br />machines (using libgfapi).<br /><br />I have a replicated+distributed setup of 4 bricks (2x2). The cluster<br />contains images for a couple of kvm virtual machines.<br /><br />My problem is that when I reboot
a brick containing a an image of a<br />VM, the VM will start throwing disk errors and eventually die.<br /><br />The gluster volume is made like this:<br /><br /># gluster vol info pool<br /><br />Volume Name: pool<br />Type: Distributed-Replicate<br />Volume ID: xxxxxxxxxxxxxxxxxxxx<br />Status: Started<br />Number of Bricks: 2 x 2 = 4<br />Transport-type: tcp<br />Bricks:<br />Brick1: srv10g:/data/gluster/brick<br />Brick2: srv11g:/data/gluster/brick<br />Brick3: srv12g:/data/gluster/brick<br />Brick4: srv13g:/data/gluster/brick<br />Options Reconfigured:<br />network.ping-timeout: 10<br />cluster.server-quorum-type: server<br />diagnostics.client-log-level: WARNING<br />auth.allow: 192.168.0.*,127.*<br />nfs.disable: on<br /><br />The KVM instances run on the same gluster bricks, with disks mounted<br />as : file=gluster://localhost/pool/images/vm-xxx-disk-1.raw,.......,cache=writethrough,aio=native<br /><br />My self-heal backlog is not always 0. It looks like some writes
are<br />not going to all bricks at the same time (?).<br /><br />gluster vol heal pool info<br /><br />sometime shows the images needing sync on one brick, the other or both.<br /><br />There are no network problems or errors on the wire.<br /><br />Any ideas what could be causing this ?<br /><br />Thanks.<br /><hr /><br />Gluster-users mailing list<br />Gluster-users@gluster.org<br /><a href="http://supercolony.gluster.org/mailman/listinfo/gluster-users">http://supercolony.gluster.org/mailman/listinfo/gluster-users</a><br /><hr /><br />Gluster-users mailing list<br />Gluster-users@gluster.org<br /><a href="http://supercolony.gluster.org/mailman/listinfo/gluster-users">http://supercolony.gluster.org/mailman/listinfo/gluster-users</a></pre></blockquote></div><br>
-- <br>
Sent from my Android device with K-9 Mail. Please excuse my brevity.</body></html>